Abstract: In classical rough set theory，the knowledge base is a partition of the universe.However，a covering is more general knowledge representation than a partition in real problems.In order to widen the application field，it is necessary to extent classical rough set theory to covering approximation space.It is a key issue for acquiring knowledge from covering approximation space to approximate a concept in it.In order to approximate crisp and fuzzy concepts，different operators are developed.However，they are not consistent.In order to make them consistent，two new operators are developed，and they are consistent with the old operators respectively，which provide a new way for the study of covering approximation operators.

Key words: covering, rough set, fuzzy concept, crisp concept, consistency

摘要： 经典粗糙集理论知识的表现形式为论域上的划分，覆盖是比划分更一般的知识表现形式。为了扩展粗糙集理论的应用领域，有必要将粗糙集理论扩展到覆盖近似空间。覆盖近似空间下的概念近似是基于覆盖近似空间知识获取的关键。针对精确概念和模糊概念，研究者定义了不同的近似方法。通过对当前的近似算子进行研究，发现了它们的不一致，并从两个角度对近似算子的定义进行了修正，从而使得它们分别与原有的算子保持一致。所得结论为覆盖近似空间下的概念近似提供了新的研究途径。

关键词: 覆盖, 粗糙集, 模糊概念, 精确概念, 一致性